Using a paired plot design to differentiate between the belowground causes and consequences of Rhamnus cathartica L. invasions

Background/Question/Methods

Using the invasive shrub Rhamnus cathartica L. (European buckthorn; hereafter buckthorn) as a model species, we developed a repeatable study framework that is capable of determining if the differences observed between soils with and without exotic shrubs pre-date invasion (possibly promoting establishment) or are shrub-induced. Ninety-seven buckthorn-invaded plots located throughout Chicago-area woodlands that varied in invasion severity were each paired to an adjacent buckthorn-free plot. Invasion severity was quantified as an invaded plot’s score along the three major axes of a PCoA ordination constructed from measures of buckthorn basal area, percent cover, height, sapling density, and invasion age (3-62 yr) made at that plot. Soil parameters hypothesized to promote and/or respond to buckthorn invasions (moisture, pH, total carbon, total nitrogen, inorganic nitrogen, calcium, leaf-litter mass, earthworm biomass, microbial community structure, and functional genes associated with nitrogen cycling) were measured at each plot. The difference in values for these RVs between each invaded and uninvaded plot of all plot pairs was calculated (ΔRVs). A relationship (linear or non-linear) between a ΔRV and severity, i.e., slope ≠ 0, indicates a buckthorn-induced change. An intercept (value of ΔRV where invasion severity = 0) different from zero indicates a pre-existing difference.

Results/Conclusions

Our framework (paired-plot design and statistical modeling of ΔRVs), unlike simply comparing differences between soils with and without exotic shrubs, can detect if these differences pre-date invasion or are shrub-induced; and unlike short-term manipulative experiments, can detect belowground changes that occur slowly over time. We found that buckthorn-invaded plots had higher levels of soil moisture, pH, total carbon, total nitrogen, ammonium-N, calcium, and Archaeal amoA genes (associated with nitrification), as well as differing fungal communities, than buckthorn-free plots prior to invasion (intercepts > 0), and that these RVs did not change as invasion severity increased (slopes = 0) as is hypothesized. Buckthorn is therefore more likely to invade where these RVs are elevated. We did, however, find that buckthorn establishes in areas with greater rates of leaf-litter decomposition (intercepts ≠ 0) and then proceeds to decrease leaf-litter mass further, and to alter levels of soil NO₃^--N and bacterial community structure (slopes ≠ 0). These findings can inform future investigations. For example, knowing that the difference in a particular RV pre-dates invasion justifies studying if this RV directly promotes invasion; while knowing that the difference in a particular RV is buckthorn-induced justifies studying the longevity and consequences of this change.